Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

Ted Cruz

7 (6.5%)

Rick Santorum

16 (14.8%)

Michelle Bachmann

13 (12%)

Chris Christie

23 (21.3%)

Rand Paul

49 (45.4%)

Total Members Voted: 107

[Criptfeind]

Alright, if you're going to change your argument from "Obviously favor some group or another" to "Not totally 100% unbiased" I'm going to switch mine "totally unbiased" (which I can see could potentially be wrong depending on what someone means by biased and I'm certainly not going to get into a semantic argument) to "almost totally unbiased" as well. And suddenly we are agreeing with each other that a program taking care of districting would be a good idea yes?

[Flarp]

You can just make a program that splits states into perfect districts in a totally non biased way.

Go for it! In plain English very high level pseudocode, in the thread!
I don't think this is as trivial as you think.

It's probably not, but for the sake of trying:

• Let n be the population of the state. Let x be the number of congressional districts to be apportioned.
• Through your favourite algorithm, assemble all possible sets of x districts of ~(n/x) size, as a combination of all secondary-level administrative divisions (usually counties, sometimes parishes).
• Taking various sundry socioeconomic markers that you wish to remain consistent (or, alternatively, diverse) throughout the state's districts, select from your combinations the set of x districts that most embody the characteristics you wish you promote in a voting population.

I suppose this isn't "unbiased" so much as "parametrized", but I'm no philosopher.

[GavJ]

Well yeah, everybody's gonna have a different "favorite algorithm" and different preferences for those socioeconomic markers, is the point.

Anyway, I dunno, yeah I guess I could agree that it's probably a good idea sight-unseen. Still worth discussing slightly more particulars though if anybody is actually interested. After all, why improve the situation 20% with a shoddy algorithm (that's still better than now), while you could improve it 60% with a damn good algorithm?

[Frumple]

Because the 20% one probably takes less effort, so you can go ahead and do that while you're messing around with the 60% one. Or the 21-59% ones. Continuous incremental improvement! It's a major business practice nowadays, might as well apply it. It can get stuff done. And in the mean time, stuff is better off.

[Criptfeind]

Okay well. I mean, Flarps suggestion is specifically tailored to be biased in certain ways, which isn't precisely wrong, but not really what I was thinking of. I was thinking more like say, The shortest-splitline method, which I believe is something that I originally found though these forums, and although you're right, I'm not a programer or a mathematician or a political 'scientist', but it well impressed me.

[GavJ]

That's not terrible. It does look like it would incur some notable overhead though by splitting weirdly in between existing government structures, requiring a bunch of strangely overlapping and redundant bureaucratic offices.

I'd suggest something like that algorithm, but as a preparatory step prior to the algorithm as stated:
0) Take any distinct towns of population smaller than, say, one tenth of a congressional district and make them nodules that the algorithm later attempts to never split apart (lines go around the edges when encountered), until/if it becomes necessary near the end of the recursion to split one or two of them.

These units would generally be too small to use the definition of those towns themselves as a source of significantly causing indirect gerrymandering, but still it would help prevent those huge state lines from cutting 25 town in half, each, creating silly extra tax dollar spending and logistics problems. Bigger population units are too much of a threat to treat this way, but also have more resources to deal with the headache anyway, so it's fine.

[Descan]

http://en.wikipedia.org/wiki/National_Popular_Vote_Interstate_Compact

[mainiac]

You could pick the electors at random from the voters and let them decide and that would give you an electoral college that was less of a factor then the current system.

Oh believe me when I say it fucks things up, but my fundamental premise is that any effort to change it (and there have been for quite a while) would require huge support that just does not exist.

Well the support does exist, it would take a majority vote in as few as five more state legislatures...

http://en.wikipedia.org/wiki/National_Popular_Vote_Interstate_Compact

[GlyphGryph]

Mine:
Take a list of every member of the state (roughly)

Randomly assign them to districts.

Done. :V

[smjjames]

Obviously the best way to do it is to take a grid and overlay it on a map of the US. Of course though, the size of the grid would be argued and the population isn't evenly distributed.

[Descan]

I wonder how Canada does it. We don't really have gerrymandering, but I don't think it's algorithmic.

Edit: Independent agencies.

[Zangi]

I wonder how Canada does it. We don't really have gerrymandering, but I don't think it's algorithmic.

Edit: Independent agencies.

Absolutely no funding/parties/donations from party affiliates right?

[Helgoland]

Germany does it by removing the incentive - if candidates from the parties' lists are added to the parliament until the seat distribution is equal to the vote distribution, getting the local candidate isn't all that important.

[Mictlantecuhtli]

Was it this topic where we discussed the Detroit mayor cutting off power to neighborhoods to help encourage them to leave their homes?

Well; here's another chapter in the saga. Now they're shutting off water for people, for reasons I don't really understand other than punishing people.

[10ebbor10]

Actually, if I recall correctly the electricity shut-off was a news fact turned propaganda*. I must assume that this is the same.

*Ie, Mayor warning: If we do not reduce power use, the grid will fail. People not reducing power => Grid fails.
Newspaper: Mayor shut down power to punish people.
Fact: Grid failed because people failing to head advice.